Sam, you're talking about absolute polarity and on some material the effect is very audible, depending quite a lot on speaker design. Some percussion and choral works highlight this effect pretty strongly. For the last 20yrs I have included facilities for selecting the output phase in most of my line pre-amplifier designs.
These days I use transformer coupled line stages so phase selection can be easily achieved with a single dpdt switch in the output. My older designs used variations on LTP topology to accomplish the same thing.
Most of the time I am a casual listener and find arbitrary phase acceptable on most material - once in a while I while discover something that just sounds wrong.
The effect is not audible on a lot of material, particularly non-acoustic music..
I found the speaker lead swapping routine annoying when I was much younger, which drove the design approach I currently use.
Whether or not this is controversial in some circles I spent a lot of time when I had the time for these things investigating the issue. Even going as far as to do blind testing to determine whether a captive audience could identify the effect, and those who were sensitive to the effect seemed in many cases to get things right about 80% or more of the time. I used popless switching, levels matched to better than 0.1dB phase to phase, arbitrary phase with switch /no switch scenarios. It was a bit complicated.. I ran about 7 people through the process, IIRC only 1 could hear no difference, and 1 was 100% correct in repeated trials. I was relatively convinced at the time that no one was keying on anything else I was doing, but you never know, and I investigated no further.
These days I use transformer coupled line stages so phase selection can be easily achieved with a single dpdt switch in the output. My older designs used variations on LTP topology to accomplish the same thing.
Most of the time I am a casual listener and find arbitrary phase acceptable on most material - once in a while I while discover something that just sounds wrong.
The effect is not audible on a lot of material, particularly non-acoustic music..
I found the speaker lead swapping routine annoying when I was much younger, which drove the design approach I currently use.
Whether or not this is controversial in some circles I spent a lot of time when I had the time for these things investigating the issue. Even going as far as to do blind testing to determine whether a captive audience could identify the effect, and those who were sensitive to the effect seemed in many cases to get things right about 80% or more of the time. I used popless switching, levels matched to better than 0.1dB phase to phase, arbitrary phase with switch /no switch scenarios. It was a bit complicated.. I ran about 7 people through the process, IIRC only 1 could hear no difference, and 1 was 100% correct in repeated trials. I was relatively convinced at the time that no one was keying on anything else I was doing, but you never know, and I investigated no further.
When my daughter was in high school I did a lot of recording of her (drums) and her friends playing musical instruments using a computer based recording system. My monitor speakers were the same Yamaha NS-10M's that I use now. The amplifiers were SS then though. I found that the persussive sounds sounded more realistic when the cone moved outward when the drum head was hit. Think of this as absolute phase from the original source to the speaker cone.
Was this observation due to speaker nonlinearities, or psychoacoustics, since I knew that the bass drum head moved toward me when she stomped the pedal? I don't know. Phase was flipped by swapping the speaker leads.
Was this observation due to speaker nonlinearities, or psychoacoustics, since I knew that the bass drum head moved toward me when she stomped the pedal? I don't know. Phase was flipped by swapping the speaker leads.
Given that there is no standard for abs phase when recording and that in those recordings well enuff done for abs phase to make a difference, probably 1/2 go one way, and the half go the other way, worrying about whether a device reverses phase seems silly -- you need to have a switch so you can change abs phase record to record (or even song to song)
dave
dave
planet10 said:
With most modern multi-track recordings absolute polarities are mixed among the many tracks anyway, even within the multi-miking of one drum kit. So the resultant mixed recording has no reference phase. For an electric guitar played through effects, how would one define correct absolute phase anyway? Nevertheless, with some multi-track recordings flipping phase DOES makes a difference, to me and to many of my audiophile friends. I suspect we're often keying on the big wavefront of the kick drum and maybe the toms, as Tubelab suggests. Sometimes one way sounds better, sometimes flipping phase just sounds different, not necessarily better or worse. Sometimes (often) there seems to be little difference at all.
But then there are some minimally-miked orchestral and small ensemble recordings that do seem to preserve a consistent phase, but it can be inverted or non-inverted, so a phase reversal switch can be useful. I have one Nonesuch LP of a percussion music ensemble that sounds dramatically better in one position than the other.
planet10 said:
Some listeners find that setting the signal polarity for recordings is useful. I wouldn't call that "silly". I suppose you mean that it doesn't much matter if any one component reverses phase because in the end you will have to try both total-system phase settings for each recording. For most of my listening, I just leave my system set to non-inverting polarity and I don't worry about it much.
Yeah, I actually caught myself while writing that but figured I'd better stick with commonly used terminology. I don't get too bent when I hear someone use the word "phase" in this context because, indeed, sine-wave phase IS shifted by 180 degrees at all frequencies when polarity is reversed relative to the non-reversed case. I take some "poetic" license with terminology occasionally, as you may have noticed...
PS: Since I believe that much of the perception of absolute polarity (there, I said it right) has to do with percussive wavefronts in the lower end of the audio band, I think loudspeaker design greatly affects the audibility of absolute polarity. For example, a ported design is a fourth-order high-pass filter. From very low frequencies to above the resonance region, there will 360 degrees of phase change relative to the input signal. Unless the speaker is very large and these phenomena are set to occur at very, very low frequencies, this variability of phase response in the bass region will muddy absolute polarity discernment. A sealed bass alignment is a second-order high-pass filter with 180 degrees of phase variation across the low end; better, but still enough to cloud things, potentially. I use Quad ESL-63s with sealed DIY subwoofers electronically equalized to fo=16 Hz and Q=0.6. I use an asymmetrical cross-over that suffers less phase irregularity in the summed response than most crossovers. As a result, I can hear polarity reversal fairly readily on some recordings. I wouldn't be surprised if I couldn't hear polarity reversal if I used a small two-way ported design.
SY said:The problem is, as we've discussed to deathD ), with defining phase for things like impulses. Or tone bursts. Or, god help me, music.
Yeah, true. "Phase" comes readily to mind to the engineer because we tend to think in terms of system frequency response, both phase response and amplitude response. Of course, the frequency response of a polarity-inverting system has a phase response of 180 degrees at all frequencies. Since any complex waveform can be devolved into sinusoidal components, we can see that each component will undergo 180 degrees of phase change though an inverting system. We can even think of an impulse in terms of its (infinite) single-frequency components, each of which will see a phase change of 180 degrees through an inverting system. The result is, of course, a flipped-polarity impulse. All of this is just my way of explaining why I don't usually wag my finger when I hear people (often engineers!) loosely use the word "phase" when they should more clearly say "polarity". I hang around engineering buddies a lot (does it show?), and we're pretty loose with terminology at times, but we all know what we mean. This can be confusing to a non-engineer listening in, and in forums like this, I (we) need to be more careful. One common mis-usage of terminology of which I've been guilty (and SY has mentioned) is referring to the quiescent or operating current of a gain device (plate/drain/collector) as the "bias current". This has been common form among my engineering cohorts for years. What we really mean of course is the "quiescent current as a result of a bias on the grid/gate/base." Since bias current more properly refers to grid/gate/base current (usually base), this can be confusing. I will accept gentle kicks in the shin whenever I get too sloppy.
planet10 said:
While that is quite true I find the switch quite useful at times, and have marked specific cds and lps where the effect is clearly audible with a recommended polarity setting. On most multi-miked recordings the effect is not always pronounced, but in some cases certain drums or vocals sound better (or as Brian remarked: different) with one polarity as opposed to the other.
As I mentioned in my original post on much of the material I listen to the polarity can be arbitrary (I used phase last time, oops. I know better, but like many I use the term interchangeably at times.) and I notice no real difference.
Another drummer joke...
A student (drummer) once pointed out to me that it is common to mike up a snare drum from both sides (perhaps with different types of microphone), then adjust the relative balance to obtain the desired sound. Given that the polarity is opposite between the two sides, which is the "correct" microphone? I wonder if it's more to do with distortion in loudspeakers.
Aircraft passenger #1: Hey I've got a really good drummer joke!
Aircraft passenger #2 (forcefully): "I'm a drummer..."
Aircraft passenger #1: "Oh, that's all right; I'll tell it really slowly."
A student (drummer) once pointed out to me that it is common to mike up a snare drum from both sides (perhaps with different types of microphone), then adjust the relative balance to obtain the desired sound. Given that the polarity is opposite between the two sides, which is the "correct" microphone? I wonder if it's more to do with distortion in loudspeakers.
Aircraft passenger #1: Hey I've got a really good drummer joke!
Aircraft passenger #2 (forcefully): "I'm a drummer..."
Aircraft passenger #1: "Oh, that's all right; I'll tell it really slowly."
I have a couple of test CD's, stereophile and a Chesky one I think, that have absolute polarity tests. I listened to the two tests. One was pretty obvious I felt while the other wasn't, to me at least. Both were acoustic music. I think the obvious one had vocals and it just seemed really closed in and off, for lack of a better word. When the polarity switched it seemed normal again.
The one that wasn't obvious to me was a jazz ensemble with the lead harmony played by a trumpet. I was a trumpet player for a number of years and despite that I couldn't really pick up which was the right polarity and the difference was pretty subtle I felt.
The obvious track didn't have any major percussive moments that I can recall.
The one that wasn't obvious to me was a jazz ensemble with the lead harmony played by a trumpet. I was a trumpet player for a number of years and despite that I couldn't really pick up which was the right polarity and the difference was pretty subtle I felt.
The obvious track didn't have any major percussive moments that I can recall.
Re: Another drummer joke...
Most good mixing boards allow you to change the output polarity on each individual microphone so that you can mix them so that they combine constructively rather than destructively canceling - particularly important when close multi-miking drum kits and the like. I've had some pretty odd effects when I forgot to do this.. (My experience is pretty limited btw, a little house PA and so recording in the dark ages so I claim no real expertise.)
EC8010 said:
Most good mixing boards allow you to change the output polarity on each individual microphone so that you can mix them so that they combine constructively rather than destructively canceling - particularly important when close multi-miking drum kits and the like. I've had some pretty odd effects when I forgot to do this..
(My experience is pretty limited btw, a little house PA and so recording in the dark ages so I claim no real expertise.)- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.